Module completion checklist

Introducing Iris flower dataset

• We are going to explore a new data set called Iris from plotly express package
• The Iris flower dataset is a multivariate data set introduced by the British statistician and biologist Ronald Fisher in his 1936 paper The use of multiple measurements in taxonomic problems.
• The data set consists of 50 samples from each of three species of Iris (Iris Setosa, Iris virginica, and Iris versicolor).
• Four features were measured from each sample: the length and the width of the sepals and petals (in centimeters)

Introducing Iris flower dataset (cont’d)

• The dataset contains a set of 150 records under 5 attributes

  • Petal Length
  • Petal Width
  • Sepal Length
  • Sepal width and
  • Class(Species)

Load the data

• Follow the steps below to read data from plotly express:

# Load the iris dataset from `plotly express`
iris_dataset = px.data.iris()

# Top 5 entries of the dataset
iris_dataset.head()

   sepal_length  sepal_width  petal_length  petal_width species  species_id
0           5.1          3.5           1.4          0.2  setosa           1
1           4.9          3.0           1.4          0.2  setosa           1
2           4.7          3.2           1.3          0.2  setosa           1
3           4.6          3.1           1.5          0.2  setosa           1
4           5.0          3.6           1.4          0.2  setosa           1

Bivariate plots

• A bivariate plot is a plot that allows us to identify the relationship between two variables

• plotly express has various bivariate plots available:

  • Scatter plots
  • Line plots
  • Funnel plots
  • Area plots

• We are going to discuss a subset of these

• For each of these plots, we will be using in-built datasets from the plotly express package

Scatter plot

• One of the most used ways to discern a relationship between two variables is to create a scatter plot
• We can create scatter plot using one variable for the x-axis and one variable for the y-axis from the dataset
• With interactive plots, we can add extra layers by changing the size and color of certain items based on other variables
• We can easily zoom in and toggle an item to make observations without creating multiple graphs

Scatter plot (cont’d)

# Construct a scatter plot by simply giving the plotting function scatter
fig = px.scatter(iris_dataset, 
          x='sepal_length', 
          y='petal_length', 
          color='species', 
          size='sepal_width')
fig.show()

Scatter plot (cont’d)

• If you have a second item you wish to toggle, try using the symbol argument

fig = px.scatter(iris_dataset, 
          x='sepal_length', 
          y='petal_length', 
          color='species',        
          size='sepal_width', 
          symbol='species')
fig.show()

Scatter plot with a linear model

fig = px.scatter(iris_dataset, 
          x='sepal_length', 
          y='petal_length', 
          color='species', 
          trendline="ols")
fig.show()

Scatter plot with a linear model (cont’d)

• We can also access the model summary from our graph to see how well the model fits our data

# Create and save model summary
results = px.get_trendline_results(fig)

# Access the model parameters
results.query("species=='setosa'").px_fit_results.iloc[0].summary()

• We’ll see the result in next slide

Scatter plot with a linear model (cont’d)

Line plot

# Create a line graph by using the plotly function `line`
fig = px.line(iris_dataset, 
          y='sepal_width',  
          color='species')
fig.show()

Multivariate plots

• Instead of just checking the relationship between two variables, maybe we want to look at the relationship between multiple pairs of variables. We can do that with plotly express as well

• Here are the types available:

  • Scatter Matrix plot
  • Parallel Coordinate plot
  • Parallel Category plot

• We’ll be discussing scatter matrix plot in next slides

Scatter matrix plot

# Construct a scatter plot by using plotly function `scatter_matrix`
fig = px.scatter_matrix(iris_dataset, 
          dimensions=["sepal_width", "sepal_length", 
                "petal_width", "petal_length"], 
          color="species")
fig.show()

Module completion checklist

Saving plots in plotly to disk

• To save plots as HTML in plotly express, we simply need the variable we saved the plot to

# Save the plot as HTML
fig.write_html('scattermatrix.html')

• If you do save it without giving the full path, make sure you change your working directory so you know where the plot is saved to

Knowledge check

Module completion checklist

Congratulations on completing this module!